Data from our own efforts and those of others indicate that different physicians discuss different things when recommending cancer screening. The U.S. Preventive Services Task Force (USPSTF) while not endorsing a specific style of physician-patient interaction, recently advocated for the use of shared decision-making when making preventive service recommendations to individual patients. We propose to use a mixed-method approach that includes both qualitative and quantitative data collection and analyses to understand the use and utility of different aspects of shared decision-making when physicians recommend CRC screening in primary care. First, we will use direct observation and audio-recording of routine health maintenance visits (n=900) to characterize the nature and content of CRC screening discussions in primary care (Aim 1). Results from these qualitative efforts will be used to derive quantitative measures characterizing physician-patient discussions of CRC screening in primary care. In the second stage, we will join these qualitatively derived measures characterizing physician-patient discussions of CRC screening with automated claims/laboratory data to determine their relationship to 12-month post-visit CRC screening use (Aim 2). A combined pre- and post-visit patient survey will allow us to assess the concordance of patient preferences for screening modality, information, and participation in decision-making with what occurs in the observed visit. We will then be able to evaluate the effect of these concordances on 12-month post-visit CRC screening use (Aim 3). We will use the Henry Ford Health System (HFHS), a large integrated delivery system serving Detroit and its surrounding suburbs, as the setting for these efforts. The diversity and size of the HFHS primary care physician and patient populations combined with the comprehensive, existing automated data systems makes this an ideal environment in which to conduct the proposed work. By joining detailed accounts of the contribution of both physicians and patients to CRC screening discussions in the real world of primary care with subsequent CRC screening use and patient reported preferences, results from the proposed project will inform a new generation of interventions aimed at improving CRC screening participation. They will provide valuable information for training both physicians and patients how to effectively discuss CRC screening as well as facilitate the development of decision aides and policies for diverse populations.